Files
pezkuwi-subxt/substrate/client/service/test/src/lib.rs
T
Aaro Altonen 80616f6d03 Integrate litep2p into Polkadot SDK (#2944)
[litep2p](https://github.com/altonen/litep2p) is a libp2p-compatible P2P
networking library. It supports all of the features of `rust-libp2p`
that are currently being utilized by Polkadot SDK.

Compared to `rust-libp2p`, `litep2p` has a quite different architecture
which is why the new `litep2p` network backend is only able to use a
little of the existing code in `sc-network`. The design has been mainly
influenced by how we'd wish to structure our networking-related code in
Polkadot SDK: independent higher-levels protocols directly communicating
with the network over links that support bidirectional backpressure. A
good example would be `NotificationHandle`/`RequestResponseHandle`
abstractions which allow, e.g., `SyncingEngine` to directly communicate
with peers to announce/request blocks.

I've tried running `polkadot --network-backend litep2p` with a few
different peer configurations and there is a noticeable reduction in
networking CPU usage. For high load (`--out-peers 200`), networking CPU
usage goes down from ~110% to ~30% (80 pp) and for normal load
(`--out-peers 40`), the usage goes down from ~55% to ~18% (37 pp).

These should not be taken as final numbers because:

a) there are still some low-hanging optimization fruits, such as
enabling [receive window
auto-tuning](https://github.com/libp2p/rust-yamux/pull/176), integrating
`Peerset` more closely with `litep2p` or improving memory usage of the
WebSocket transport
b) fixing bugs/instabilities that incorrectly cause `litep2p` to do less
work will increase the networking CPU usage
c) verification in a more diverse set of tests/conditions is needed

Nevertheless, these numbers should give an early estimate for CPU usage
of the new networking backend.

This PR consists of three separate changes:
* introduce a generic `PeerId` (wrapper around `Multihash`) so that we
don't have use `NetworkService::PeerId` in every part of the code that
uses a `PeerId`
* introduce `NetworkBackend` trait, implement it for the libp2p network
stack and make Polkadot SDK generic over `NetworkBackend`
  * implement `NetworkBackend` for litep2p

The new library should be considered experimental which is why
`rust-libp2p` will remain as the default option for the time being. This
PR currently depends on the master branch of `litep2p` but I'll cut a
new release for the library once all review comments have been
addresses.

---------

Signed-off-by: Alexandru Vasile <alexandru.vasile@parity.io>
Co-authored-by: Dmitry Markin <dmitry@markin.tech>
Co-authored-by: Alexandru Vasile <60601340+lexnv@users.noreply.github.com>
Co-authored-by: Alexandru Vasile <alexandru.vasile@parity.io>
2024-04-08 16:44:13 +00:00

574 lines
17 KiB
Rust

// This file is part of Substrate.
// Copyright (C) Parity Technologies (UK) Ltd.
// SPDX-License-Identifier: GPL-3.0-or-later WITH Classpath-exception-2.0
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
// You should have received a copy of the GNU General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
//! Service integration test utils.
use futures::{task::Poll, Future, TryFutureExt as _};
use log::{debug, info};
use parking_lot::Mutex;
use sc_client_api::{Backend, CallExecutor};
use sc_network::{
config::{MultiaddrWithPeerId, NetworkConfiguration, TransportConfig},
multiaddr, NetworkBlock, NetworkPeers, NetworkStateInfo,
};
use sc_network_sync::SyncingService;
use sc_service::{
client::Client,
config::{BasePath, DatabaseSource, KeystoreConfig, RpcBatchRequestConfig},
BlocksPruning, ChainSpecExtension, Configuration, Error, GenericChainSpec, Role,
RuntimeGenesis, SpawnTaskHandle, TaskManager,
};
use sc_transaction_pool_api::TransactionPool;
use sp_blockchain::HeaderBackend;
use sp_runtime::traits::Block as BlockT;
use std::{iter, net::Ipv4Addr, pin::Pin, sync::Arc, task::Context, time::Duration};
use tempfile::TempDir;
use tokio::{runtime::Runtime, time};
#[cfg(test)]
mod client;
/// Maximum duration of single wait call.
const MAX_WAIT_TIME: Duration = Duration::from_secs(60 * 3);
struct TestNet<G, E, F, U> {
runtime: Runtime,
authority_nodes: Vec<(usize, F, U, MultiaddrWithPeerId)>,
full_nodes: Vec<(usize, F, U, MultiaddrWithPeerId)>,
chain_spec: GenericChainSpec<G, E>,
base_port: u16,
nodes: usize,
}
impl<G, E, F, U> Drop for TestNet<G, E, F, U> {
fn drop(&mut self) {
// Drop the nodes before dropping the runtime, as the runtime otherwise waits for all
// futures to be ended and we run into a dead lock.
self.full_nodes.drain(..);
self.authority_nodes.drain(..);
}
}
pub trait TestNetNode: Clone + Future<Output = Result<(), Error>> + Send + 'static {
type Block: BlockT;
type Backend: Backend<Self::Block>;
type Executor: CallExecutor<Self::Block> + Send + Sync;
type RuntimeApi: Send + Sync;
type TransactionPool: TransactionPool<Block = Self::Block>;
fn client(&self) -> Arc<Client<Self::Backend, Self::Executor, Self::Block, Self::RuntimeApi>>;
fn transaction_pool(&self) -> Arc<Self::TransactionPool>;
fn network(&self) -> Arc<dyn sc_network::service::traits::NetworkService>;
fn sync(&self) -> &Arc<SyncingService<Self::Block>>;
fn spawn_handle(&self) -> SpawnTaskHandle;
}
pub struct TestNetComponents<TBl: BlockT, TBackend, TExec, TRtApi, TExPool> {
task_manager: Arc<Mutex<TaskManager>>,
client: Arc<Client<TBackend, TExec, TBl, TRtApi>>,
transaction_pool: Arc<TExPool>,
network: Arc<dyn sc_network::service::traits::NetworkService>,
sync: Arc<SyncingService<TBl>>,
}
impl<TBl: BlockT, TBackend, TExec, TRtApi, TExPool>
TestNetComponents<TBl, TBackend, TExec, TRtApi, TExPool>
{
pub fn new(
task_manager: TaskManager,
client: Arc<Client<TBackend, TExec, TBl, TRtApi>>,
network: Arc<dyn sc_network::service::traits::NetworkService>,
sync: Arc<SyncingService<TBl>>,
transaction_pool: Arc<TExPool>,
) -> Self {
Self {
client,
sync,
transaction_pool,
network,
task_manager: Arc::new(Mutex::new(task_manager)),
}
}
}
impl<TBl: BlockT, TBackend, TExec, TRtApi, TExPool> Clone
for TestNetComponents<TBl, TBackend, TExec, TRtApi, TExPool>
{
fn clone(&self) -> Self {
Self {
task_manager: self.task_manager.clone(),
client: self.client.clone(),
transaction_pool: self.transaction_pool.clone(),
network: self.network.clone(),
sync: self.sync.clone(),
}
}
}
impl<TBl: BlockT, TBackend, TExec, TRtApi, TExPool> Future
for TestNetComponents<TBl, TBackend, TExec, TRtApi, TExPool>
{
type Output = Result<(), Error>;
fn poll(self: Pin<&mut Self>, cx: &mut Context) -> Poll<Self::Output> {
Pin::new(&mut self.task_manager.lock().future()).poll(cx)
}
}
impl<TBl, TBackend, TExec, TRtApi, TExPool> TestNetNode
for TestNetComponents<TBl, TBackend, TExec, TRtApi, TExPool>
where
TBl: BlockT,
TBackend: sc_client_api::Backend<TBl> + Send + Sync + 'static,
TExec: CallExecutor<TBl> + Send + Sync + 'static,
TRtApi: Send + Sync + 'static,
TExPool: TransactionPool<Block = TBl> + Send + Sync + 'static,
{
type Block = TBl;
type Backend = TBackend;
type Executor = TExec;
type RuntimeApi = TRtApi;
type TransactionPool = TExPool;
fn client(&self) -> Arc<Client<Self::Backend, Self::Executor, Self::Block, Self::RuntimeApi>> {
self.client.clone()
}
fn transaction_pool(&self) -> Arc<Self::TransactionPool> {
self.transaction_pool.clone()
}
fn network(&self) -> Arc<dyn sc_network::service::traits::NetworkService> {
self.network.clone()
}
fn sync(&self) -> &Arc<SyncingService<Self::Block>> {
&self.sync
}
fn spawn_handle(&self) -> SpawnTaskHandle {
self.task_manager.lock().spawn_handle()
}
}
impl<G, E, F, U> TestNet<G, E, F, U>
where
F: Clone + Send + 'static,
U: Clone + Send + 'static,
{
pub fn run_until_all_full<FP>(&mut self, full_predicate: FP)
where
FP: Send + Fn(usize, &F) -> bool + 'static,
{
let full_nodes = self.full_nodes.clone();
let future = async move {
let mut interval = time::interval(Duration::from_millis(100));
loop {
interval.tick().await;
if full_nodes.iter().all(|(id, service, _, _)| full_predicate(*id, service)) {
break
}
}
};
if self
.runtime
.block_on(async move { time::timeout(MAX_WAIT_TIME, future).await })
.is_err()
{
panic!("Waited for too long");
}
}
}
fn node_config<
G: RuntimeGenesis + 'static,
E: ChainSpecExtension + Clone + 'static + Send + Sync,
>(
index: usize,
spec: &GenericChainSpec<G, E>,
role: Role,
tokio_handle: tokio::runtime::Handle,
key_seed: Option<String>,
base_port: u16,
root: &TempDir,
) -> Configuration {
let root = root.path().join(format!("node-{}", index));
let mut network_config = NetworkConfiguration::new(
format!("Node {}", index),
"network/test/0.1",
Default::default(),
None,
);
network_config.allow_non_globals_in_dht = true;
network_config.listen_addresses.push(
iter::once(multiaddr::Protocol::Ip4(Ipv4Addr::new(127, 0, 0, 1)))
.chain(iter::once(multiaddr::Protocol::Tcp(base_port + index as u16)))
.collect(),
);
network_config.transport =
TransportConfig::Normal { enable_mdns: false, allow_private_ip: true };
Configuration {
impl_name: String::from("network-test-impl"),
impl_version: String::from("0.1"),
role,
tokio_handle,
transaction_pool: Default::default(),
network: network_config,
keystore: KeystoreConfig::Path { path: root.join("key"), password: None },
database: DatabaseSource::RocksDb { path: root.join("db"), cache_size: 128 },
trie_cache_maximum_size: Some(16 * 1024 * 1024),
state_pruning: Default::default(),
blocks_pruning: BlocksPruning::KeepFinalized,
chain_spec: Box::new((*spec).clone()),
wasm_method: Default::default(),
wasm_runtime_overrides: Default::default(),
rpc_addr: Default::default(),
rpc_max_connections: Default::default(),
rpc_cors: None,
rpc_methods: Default::default(),
rpc_max_request_size: Default::default(),
rpc_max_response_size: Default::default(),
rpc_id_provider: Default::default(),
rpc_max_subs_per_conn: Default::default(),
rpc_port: 9944,
rpc_message_buffer_capacity: Default::default(),
rpc_batch_config: RpcBatchRequestConfig::Unlimited,
rpc_rate_limit: None,
prometheus_config: None,
telemetry_endpoints: None,
default_heap_pages: None,
offchain_worker: Default::default(),
force_authoring: false,
disable_grandpa: false,
dev_key_seed: key_seed,
tracing_targets: None,
tracing_receiver: Default::default(),
max_runtime_instances: 8,
announce_block: true,
base_path: BasePath::new(root.clone()),
data_path: root,
informant_output_format: Default::default(),
runtime_cache_size: 2,
}
}
impl<G, E, F, U> TestNet<G, E, F, U>
where
F: TestNetNode,
E: ChainSpecExtension + Clone + 'static + Send + Sync,
G: RuntimeGenesis + 'static,
{
fn new(
temp: &TempDir,
spec: GenericChainSpec<G, E>,
full: impl Iterator<Item = impl FnOnce(Configuration) -> Result<(F, U), Error>>,
authorities: impl Iterator<Item = (String, impl FnOnce(Configuration) -> Result<(F, U), Error>)>,
base_port: u16,
) -> TestNet<G, E, F, U> {
sp_tracing::try_init_simple();
fdlimit::raise_fd_limit().unwrap();
let runtime = Runtime::new().expect("Error creating tokio runtime");
let mut net = TestNet {
runtime,
authority_nodes: Default::default(),
full_nodes: Default::default(),
chain_spec: spec,
base_port,
nodes: 0,
};
net.insert_nodes(temp, full, authorities);
net
}
fn insert_nodes(
&mut self,
temp: &TempDir,
full: impl Iterator<Item = impl FnOnce(Configuration) -> Result<(F, U), Error>>,
authorities: impl Iterator<Item = (String, impl FnOnce(Configuration) -> Result<(F, U), Error>)>,
) {
self.runtime.block_on(async {
let handle = self.runtime.handle().clone();
for (key, authority) in authorities {
let node_config = node_config(
self.nodes,
&self.chain_spec,
Role::Authority,
handle.clone(),
Some(key),
self.base_port,
temp,
);
let addr = node_config.network.listen_addresses.first().unwrap().clone();
let (service, user_data) =
authority(node_config).expect("Error creating test node service");
handle.spawn(service.clone().map_err(|_| ()));
let addr = MultiaddrWithPeerId {
multiaddr: addr,
peer_id: service.network().local_peer_id(),
};
self.authority_nodes.push((self.nodes, service, user_data, addr));
self.nodes += 1;
}
for full in full {
let node_config = node_config(
self.nodes,
&self.chain_spec,
Role::Full,
handle.clone(),
None,
self.base_port,
temp,
);
let addr = node_config.network.listen_addresses.first().unwrap().clone();
let (service, user_data) =
full(node_config).expect("Error creating test node service");
handle.spawn(service.clone().map_err(|_| ()));
let addr = MultiaddrWithPeerId {
multiaddr: addr,
peer_id: service.network().local_peer_id(),
};
self.full_nodes.push((self.nodes, service, user_data, addr));
self.nodes += 1;
}
});
}
}
fn tempdir_with_prefix(prefix: &str) -> TempDir {
tempfile::Builder::new()
.prefix(prefix)
.tempdir()
.expect("Error creating test dir")
}
pub fn connectivity<G, E, Fb, F>(spec: GenericChainSpec<G, E>, full_builder: Fb)
where
E: ChainSpecExtension + Clone + 'static + Send + Sync,
G: RuntimeGenesis + 'static,
Fb: Fn(Configuration) -> Result<F, Error>,
F: TestNetNode,
{
const NUM_FULL_NODES: usize = 5;
let expected_full_connections = NUM_FULL_NODES - 1;
{
let temp = tempdir_with_prefix("substrate-connectivity-test");
{
let mut network = TestNet::new(
&temp,
spec.clone(),
(0..NUM_FULL_NODES).map(|_| |cfg| full_builder(cfg).map(|s| (s, ()))),
// Note: this iterator is empty but we can't just use `iter::empty()`, otherwise
// the type of the closure cannot be inferred.
(0..0).map(|_| (String::new(), { |cfg| full_builder(cfg).map(|s| (s, ())) })),
30400,
);
info!("Checking star topology");
let first_address = network.full_nodes[0].3.clone();
for (_, service, _, _) in network.full_nodes.iter().skip(1) {
service
.network()
.add_reserved_peer(first_address.clone())
.expect("Error adding reserved peer");
}
network.run_until_all_full(move |_index, service| {
let connected = service.network().sync_num_connected();
debug!("Got {}/{} full connections...", connected, expected_full_connections);
connected == expected_full_connections
});
};
temp.close().expect("Error removing temp dir");
}
{
let temp = tempdir_with_prefix("substrate-connectivity-test");
{
let mut network = TestNet::new(
&temp,
spec,
(0..NUM_FULL_NODES).map(|_| |cfg| full_builder(cfg).map(|s| (s, ()))),
// Note: this iterator is empty but we can't just use `iter::empty()`, otherwise
// the type of the closure cannot be inferred.
(0..0).map(|_| (String::new(), { |cfg| full_builder(cfg).map(|s| (s, ())) })),
30400,
);
info!("Checking linked topology");
let mut address = network.full_nodes[0].3.clone();
for i in 0..NUM_FULL_NODES {
if i != 0 {
if let Some((_, service, _, node_id)) = network.full_nodes.get(i) {
service
.network()
.add_reserved_peer(address)
.expect("Error adding reserved peer");
address = node_id.clone();
}
}
}
network.run_until_all_full(move |_index, service| {
let connected = service.network().sync_num_connected();
debug!("Got {}/{} full connections...", connected, expected_full_connections);
connected == expected_full_connections
});
}
temp.close().expect("Error removing temp dir");
}
}
pub fn sync<G, E, Fb, F, B, ExF, U>(
spec: GenericChainSpec<G, E>,
full_builder: Fb,
mut make_block_and_import: B,
mut extrinsic_factory: ExF,
) where
Fb: Fn(Configuration) -> Result<(F, U), Error>,
F: TestNetNode,
B: FnMut(&F, &mut U),
ExF: FnMut(&F, &U) -> <F::Block as BlockT>::Extrinsic,
U: Clone + Send + 'static,
E: ChainSpecExtension + Clone + 'static + Send + Sync,
G: RuntimeGenesis + 'static,
{
const NUM_FULL_NODES: usize = 10;
const NUM_BLOCKS: usize = 512;
let temp = tempdir_with_prefix("substrate-sync-test");
let mut network = TestNet::new(
&temp,
spec,
(0..NUM_FULL_NODES).map(|_| |cfg| full_builder(cfg)),
// Note: this iterator is empty but we can't just use `iter::empty()`, otherwise
// the type of the closure cannot be inferred.
(0..0).map(|_| (String::new(), { |cfg| full_builder(cfg) })),
30500,
);
info!("Checking block sync");
let first_address = {
let &mut (_, ref first_service, ref mut first_user_data, _) = &mut network.full_nodes[0];
for i in 0..NUM_BLOCKS {
if i % 128 == 0 {
info!("Generating #{}", i + 1);
}
make_block_and_import(first_service, first_user_data);
}
let info = network.full_nodes[0].1.client().info();
network.full_nodes[0]
.1
.sync()
.new_best_block_imported(info.best_hash, info.best_number);
network.full_nodes[0].3.clone()
};
info!("Running sync");
for (_, service, _, _) in network.full_nodes.iter().skip(1) {
service
.network()
.add_reserved_peer(first_address.clone())
.expect("Error adding reserved peer");
}
network.run_until_all_full(|_index, service| {
service.client().info().best_number == (NUM_BLOCKS as u32).into()
});
info!("Checking extrinsic propagation");
let first_service = network.full_nodes[0].1.clone();
let first_user_data = &network.full_nodes[0].2;
let best_block = first_service.client().info().best_hash;
let extrinsic = extrinsic_factory(&first_service, first_user_data);
let source = sc_transaction_pool_api::TransactionSource::External;
futures::executor::block_on(
first_service.transaction_pool().submit_one(best_block, source, extrinsic),
)
.expect("failed to submit extrinsic");
network.run_until_all_full(|_index, service| service.transaction_pool().ready().count() == 1);
}
pub fn consensus<G, E, Fb, F>(
spec: GenericChainSpec<G, E>,
full_builder: Fb,
authorities: impl IntoIterator<Item = String>,
) where
Fb: Fn(Configuration) -> Result<F, Error>,
F: TestNetNode,
E: ChainSpecExtension + Clone + 'static + Send + Sync,
G: RuntimeGenesis + 'static,
{
const NUM_FULL_NODES: usize = 10;
const NUM_BLOCKS: usize = 10; // 10 * 2 sec block production time = ~20 seconds
let temp = tempdir_with_prefix("substrate-consensus-test");
let mut network = TestNet::new(
&temp,
spec,
(0..NUM_FULL_NODES / 2).map(|_| |cfg| full_builder(cfg).map(|s| (s, ()))),
authorities
.into_iter()
.map(|key| (key, { |cfg| full_builder(cfg).map(|s| (s, ())) })),
30600,
);
info!("Checking consensus");
let first_address = network.authority_nodes[0].3.clone();
for (_, service, _, _) in network.full_nodes.iter() {
service
.network()
.add_reserved_peer(first_address.clone())
.expect("Error adding reserved peer");
}
for (_, service, _, _) in network.authority_nodes.iter().skip(1) {
service
.network()
.add_reserved_peer(first_address.clone())
.expect("Error adding reserved peer");
}
network.run_until_all_full(|_index, service| {
service.client().info().finalized_number >= (NUM_BLOCKS as u32 / 2).into()
});
info!("Adding more peers");
network.insert_nodes(
&temp,
(0..NUM_FULL_NODES / 2).map(|_| |cfg| full_builder(cfg).map(|s| (s, ()))),
// Note: this iterator is empty but we can't just use `iter::empty()`, otherwise
// the type of the closure cannot be inferred.
(0..0).map(|_| (String::new(), { |cfg| full_builder(cfg).map(|s| (s, ())) })),
);
for (_, service, _, _) in network.full_nodes.iter() {
service
.network()
.add_reserved_peer(first_address.clone())
.expect("Error adding reserved peer");
}
network.run_until_all_full(|_index, service| {
service.client().info().finalized_number >= (NUM_BLOCKS as u32).into()
});
}